Table 5. Summary of studies assessing service utilisation.
| Authors | Country /
region |
Data used | Outcome | Findings | Comment / Limitations |
|---|---|---|---|---|---|
| Capuzzi et al., 2020 | Italy | Emergency psychiatric evaluations at
psychiatric emergency rooms in two centres in Lombardy, serving a population of approx. 850,000 in two equivalent periods pre (Fri 22 Feb 2019-Sun 5 May 2019) and following the first COVID-19 case in Italy up to end of first phase of lock-down (Fri 21 Feb 2020 to Sun 3rd May 2020). Data obtained from hospital registers. |
Suicide
attempts/self- harm |
Period A (2019) 388 total attendances,
including 68 (17.5%) for self-harm/suicide attempt Period B (2020) 225 total attendances, including 59 (26.2%) for self-harm/suicide attempt. Whilst absolute number of SH/SA cases lower, the difference in number as a proportion of total cases was somewhat higher in age/sex adjusted models (aOR 1.48 (0.97 to 2.28) |
Hospital based study from two
centres Peer reviewed |
| Chen et al., 2020 | England, UK | Data obtained from Trust hospitals clinical
record systems. People using or referred to inpatient and community MH services (including psychological therapy services) in Cambridge and Peterborough - population approx 860,000. Data for Liaison psychiatry referrals for SH/Suicide attempt/ cover 11 March 2014 - 30 August 2020. Data also presented for suicidal thoughts, but data were combined with “low mood” |
Intentional drug
overdose and self-harm |
A marked reduction (p<0.001) in liaison
psychiatry referrals for intentional drug overdose, self-harm and suicidal thoughts occurred after 23 March (lockdown). The proportion of referrals returned to pre-lockdown levels by May/June 2020. |
Liaison team referral only (not all ED
attendances) at a single hospital. Liaison psychiatry referral pathways may have changed as a result of COVID-19 No detailed demographic analysis of referrals as the paper focused on a wide range of mental and physical health presentations. Single area in England. Peer reviewed |
| Dragovic et al., 2020 | Australia | Western Australia (WA) North Metropolitan
Health Services EDs were extracted from the Emergency Department Data Collection database. These 3 EDs serve a population of approx. 800,000 persons. Attendances over the period January to May 2020 were compared to those that occurred over the same calendar month periods during 2019. |
Suicide
attempts/self- harm |
7140 attendances (5522 persons) over the
two study periods. Suicidal and self-harm presentation decreased by 26% to previous year |
Attendances at three hospitals but
WA has low population density and went into stringent lockdown early - hence findings may not be generalisable to other Australian states or other countries; routinely collected healthcare data are large and complete, but they lack rich contextual detail. Peer reviewed |
| Gonçalves-Pinho et al., 2021 | Portugal l | People attending a Psychiatric Emergency
Department in a tertiary hospital in North Portugal serving a population of approximately 3 million people. Attendance between March 19th and May 2nd 2020 (when "emergency state" / restriction of movement existed in Portugal in response to COVID-19) compared with same dates in 2019 |
“Suicide
and self- inflicted injury presentations” to psychiatric ED |
Between March 19
th and May 2
nd 2020,
a significant reduction was identified in presentations of “suicide and intentional self-inflicted injury” to a metropolitan psychiatric ED, compared to the same period in 2019: N=36 v 81, a 55.6% reduction. |
Based on attendances at a single
hospital. Unclear if codes include people with suicidal thoughts as well as acts. Peer reviewed |
| Hernández-Calle et al., 2020 | Spain | Electronic health records examined at a major
general hospital in Madrid, Spain: November 2018 to April 2020. |
Suicidal
thoughts |
During March-April 2020, significantly
fewer psychiatric emergency department visits due to suicidal ideation were reported compared to the same period in 2019. |
Data only shown in a graph.
Single centre study - findings may have limited generalisability. Peer reviewed |
| Hewson et al., 2020 | UK | 31 prisons in UK
Internal reports from Safer Custody Units in 31 prisons where healthcare is provided by CareUK (Russell Green, personal communication) |
Suicide
attempts/self- harm |
After lockdown there were fewer
implementations of Assessment, Care in Custody and Teamwork (ACCT) processes; to initiate care- plans for prisoners considered at risk of self-harm or suicide. Across the 31 prisons, there were 1079 ACCTs implemented in February 2020 compared to 828 in April 2020, a fall of just under 25%. Analysis of data for 8 prisons indicated that there were falls in incidents of self-harm, decreasing by a third from 324 in February 2020 to 214 in April 2020. |
No gender breakdown (female
prisoners in the UK generally have much higher rates of self-harm than male prisoners) Unclear the basis of the selection of the 8 prisons with self-harm data Peer reviewed editorial |
| Jacob et al., 2020 | Australia | Single trauma centre in Australia, serving a
population of 1.5 million. Compared mean number of trauma admissions during March and April during years 2016 to 2020 |
Self-harm | During March and April 2020 a significant
decrease in total number of trauma- related admissions was observed, but no significant difference in admissions following self-harm was seen. |
Mean no. of admissions examined
before and during the Covid-19 public health emergency. Findings from a single centre may not be generalisable. The study was evidently under-powered for examination of mean monthly self- harm admissions. Peer reviewed |
| Karakasi et al., 2020a | Greece | Records of psychiatric emergency cases
presenting at the psychiatric emergency department of AHEPA University General Hospital of Thessaloniki during the following equal time intervals: 1 March to 15-May 2019, 15November 2019 to 31 January 2020, 1 March to 15 May 2020. |
Suicide
attempts/self- harm |
During the restrictive measures in Greece
(March – May 2020), the number of hospital presenting emergency psychiatric incidents fell by half (p < 0.01). The number of suicide attempts was higher in March- May 2020 (n=7) compared to the same period in 2019 (n= 5) and Nov 2019-Jan 2020 (n=4) |
Data from a single hospital
Small numbers Letter. Uncertain if peer-reviewed |
| Lersch, 2020 | USA | Emergency calls (911) to Detroit Police
Department for services between 26th Feb (first reported case COVID in city) and 27th April 2020. Comparison with 2017–2019 and also number of COVID-19 cases in the city. |
Suicide threats
and suicides in progress |
In the time period of interest during the
pandemic in 2020, the number of 911 calls for mental health issues was the lowest of the 4 years (2017–2020), declining by 16% from 2019 to 2020. However, the number of calls for suicide threats declined in 2020, while the number of calls for suicides in progress remained relatively stable over the 4-year period. No significant correlations between daily number of COVID-19 cases in the city and the number of calls from mentally ill persons, but as the number of COVID-19 cases increased there was a decline in calls for suicides in progress, but a significant inverse correlation between numbers of COVID-19 cases and threats of suicide calls (Pearson’s r=0.394) and a similar but non-significant relationship with calls for suicides in progress. In local area analysis, “some of the ‘hotspots’ for suicide threats were in areas of higher rates of COVID-19 cases”. |
Interesting analysis by numbers
of COVID-19 cases, including by locality. Single city. Data are early and may not be complete for COVID-19 cases. Unclear if peer-reviewed |
| McAndrew et al., 2020 | Ireland | Electronic health records for the emergency
department (ED) of a large teaching hospital in Dublin were examined during the first 8 weeks of the Covid-19 emergency (from 16th March to 10th May 2020). Comparative data for 2018 and 2019 were also examined. |
Suicide
attempts/self- harm, Suicidal thoughts |
A 21% reduction in the frequency of
psychiatric emergency presentations was observed, although the proportion of presentations with suicidal ideation or self-harm as factors remained unchanged. The observed reduction was largely due to a reduce attendance frequency during 'normal' hours. |
Electronic health record studies are
not prone to selection or self-report information biases. Further research examining patterns of emergency psychiatry presentations during COVID-19 could identify risky / vulnerable groups of people who have not been seeking help during a crisis. Similar studies from other countries and with extended follow-up periods are needed to build up a comprehensive picture of these temporal patterns. Peer reviewed |
| McIntyre et al., 2020 | Ireland | Self-harm referrals to Liaison Psychiatry team
in a single tertiary care hospital in Gallway Ireland. Contrast 1 March 2020–31 May 2020 with the same period in 2017–2019 |
Self-harm
presentations to a general hospital. |
Between March-April 2020, a significantly
lower proportion of self-harm presentations (-35%) to the hospital was reported, compared to the same period for 2017–2019. At the end of May, similar proportions of self-harm presentations were reported compared to previous years. |
Single hospital study. Incidence
based on referrals to liaison psychiatry - may under-estimate total hospital presenting cases. Liaison psychiatry referral pathways may have changed as a result of COVID-19. Peer reviewed |
| Olding et al., 2021 | England, UK | Trauma patients with penetrating injuries
who were treated at King's College Hospital in London, 23rd March to 29th April 2020 compared to the same period in 2018 and 2019. |
Self-harm (self-
inflicted injuries |
Whilst the incidence of all types of
penetrating trauma appeared to have fallen by 35% during the early lockdown period), the number of self-harm episodes increased from n=1 in 2018 to 5 in 2019 and 8 in 2020 |
Small, single site study. Crude
analytical approach. Number of self- harm cases too small to draw any strong conclusions Peer reviewed |
| Pignon et al., 2020a | France | Emergency psychiatric consultations from
three psychiatric emergency centres from first four weeks of lockdown (started March 17th 2020) and corresponding weeks 2019 |
Suicide
attempts |
During the four first weeks of lockdown,
553 emergency psychiatric consultations were carried out, less than half (45.2%) of the corresponding weeks in 2019 (1224 consultations). Total suicide attempts decreased in 2020 to 42.6% of those in 2019. |
Peer reviewed publication now
published, Pignon et al., 2020b |
| Rajput et al., 2020 | England, UK | Trauma admissions to a single level 1 trauma
centre in Liverpool using data from a trauma research network database. Compared three 7-week periods: (1) Lockdown: 23 March 2020–10 May 2020) (2) Pre-lockdown: 7 weeks prior to lockdown (27 January 2020–15 March 2020) (3) Pre-lockdown 2019: 7 week equivalent period in 2019 (25 March 2019–12 May 2019) |
Suicide
attempts/self- harm |
Total trauma centre attendances fell
during lockdown: 2019: n=194; 7 weeks pre lockdown 2020 n=173; during lockdown n=121 Equivalent numbers for self-harm were: 20 (2019); 24 (pre-lockdown 2020); 14 (lockdown 2020): i.e. 30% fall vs 2019. |
Small sample size; no assessment of
any change in socio-demographic characteristics of self-harm; possible changes due to service re-configurations in response to COVID. Peer reviewed |
| Rhodes et al., 2020 | USA | Trauma registry data of attendees at a Level 1
trauma centre in S Carolina, USA Jan 1-May 1 2019 compared to Jan 1 - May 1 2020 (lockdown April 8 th – May 1 st 2020). |
Suicide
attempts and self-harm, including specific methods |
Some evidence of rise in suicide attempts:
2019: 6 (0.6% of all presentations); 2020: 11 (1.4%) (p=0.079), including ‘self- harm by jumping’: 2019: 0 (0%); 2020: 5 (0.6%); p=0.011). No change in other ‘self-harm’ presentations: gun: 2019: 4 (0.4%); 2020: 4 (0.5%) (p=0.716); knife: 2019: 2 (0.2%); 2020: 1 (0.1%) (p=0.719), nor in acts of ‘Undetermined intent’: 2019: 18 (1.8%); 2020: 6 (0.8%) (p=0.064). |
Most of the period studied (15 of
the 18 weeks) in 2020 preceded lockdown. Small numbers and no specific data on suicide attempts during the post-lockdown period. The statistical comparison of suicide/SH episodes compared these episodes as a % of total attendances, rather than changes in absolute numbers. Peer reviewed |
| Sade et al., 2020 | Israel | Pregnant women admitted to high risk
pregnancy units between 19 March 2020 and 26 May 2020 (the strict isolation period of the pandemic) (n=90) compared to those hospitalised to these units between November 2016 and April 2017 (n=279) |
Suicidal
thoughts assessed using the Edinburgh postnatal depression scale (EPDS) |
Prevalence of suicidal thoughts was similar
pre (5.0%) vs during (8.6%) pandemic (p = 0.221). OR in multivariable logistic regression model, controlling for maternal age, adjusted OR 1.8, 95% CI 0.71–4.85, p = 0.203. |
Admission criteria may have
changed post pandemic (although admissions per month similar ~ 45/month Relatively small sample Select sample - pregnant women - generalisability to wider population uncertain. Pre-pandemic data collected in Nov 2016-April 2017 - 3 years previously - no account of any secular trends (also seasonal difference in collection period). Peer reviewed |
| Sheridan et al., 2021 | USA | Emergency department visits for mental
health issues to a single tertiary care pediatric hospital in Portland, Oregon April 1 2019 up to 29 April 2020 |
Suicidal patients | Department dealt with 14108 patients in
2019. 16 suicidal patients seen in April 2020 vs. 46 in April 2020 (a 65% fall) |
Before / after lock down
comparison, time trend analysis Used routinely available data Data on suicidal patients only specified for 1 month. One tertiary centre so not generalisable. Peer-reviewed |
| Smalley et al., 2021 | USA | Attendees with suicidal thoughts and alcohol
issues across 20 diverse EDs in a large Midwest integrated healthcare system with >750,000 ED visits annually. All behavioural health (BH) visits were collected for 1-month (March 25 th to April 24, 2020) following “stay at home” orders (lockdown). ICD-10 codes were used to identify visits associated with suicidal thoughts. The same parameters were used to collect data for the same time period for 2019. |
Suicidal
thoughts ICD coded by hospital staff |
Comparing 2020 with the same period in
2019, there was 44.4% decrease in overall ED visits and 28.0% decrease in BH visits. Attendances of individuals with suicidal thoughts decreased by 60.6% in 2020 (n=451) vs. 2019 (n=1144). As a percentage of all ED attendances, suicidal thoughts attendances decreased from 2.03% in 2019 to 1.44% in 2020. |
Alternative avenues for help-
seeking not included. But highlights importance of improving access for vulnerable populations during a pandemic. Included only one month in 2019 and one in 2020. Letter to editor, probably not peer reviewed |
| Titov et al., 2020 | Australa | Callers / website visits to "Mindspot" - national
digital MH service in Australia. Compared caller volume and characteristics 1- 28 Sept 2019 (n=1650) vs. 19 March - 15 April 2020 (n=1668) |
Suicidal
thoughts question from PHQ-9 |
No change in prevalence of: a) suicidal
thoughts (30.6% in September 2019 vs. 27.5% in March-April 2020; p=0.08), or b) suicidal intentions or plans (3.7% v 2.9% post p=0.27) |
Clinical / helpline sample - not
population based. Possible seasonal differences- September contacts vs. March-April Evidence of increased contact volume to a digital service. Peer reviewed |
| Walker et al., 2020 | USA | ED attendances (adult and pediatric) from
an integrated multiple hospital / ED system. n=18 EDs across several states. Diagnoses via electronic health records. Pandemic period (17 March 2020 to 21 April 2020) compared to same period in 2019 (17 March 2019 to 21 April 2019) and 36 day pre- pandemic period in 2020 (9 Feb 2020 to 16 March 2020) |
Suicide
attempts/self- harm |
Total ED attendances fell by around 50% during the period
of "the broad institution of distancing measures in response to the COVID-19 pandemic". Likewise, total ED attendances with "suicide" diagnosis fell by around one third during pandemic period: 17 March 2020 to 21 April 2020: n=36 (0.2% of total attendances) vs. 17 March 2019 to 21 April 2019: n=59 (0.2% total attendances) 9 Feb 2020 to 16 March vs. 2020: n=64 (0.2% total) |
Hospital presentations only
Only includes first 36 days of distancing measures. Peer reviewed |